地理科学进展 ›› 2008, Vol. 27 ›› Issue (4): 15-22.doi: 10.11820/dlkxjz.2008.04.003

• 水系统模拟研究专栏 • 上一篇    下一篇

基于静止气象卫星观测的降水时间降尺度研究

师春香1,2,3, 谢正辉1   

  1. 1. 中国科学院大气物理研究所, 北京100029;
    2. 中国气象局国家卫星气象中心, 北京100081;
    3. 中国科学院研究生院, 北京100039
  • 收稿日期:2008-04-01 修回日期:2008-06-01 出版日期:2008-07-25 发布日期:2008-07-25
  • 通讯作者: 谢正辉,男,教授.E- mail:zxie@lasg.iap.ac.cn E-mail:zxie@lasg.iap.ac.cn
  • 作者简介:师春香(1964- ), 女, 研究员, 从事卫星遥感及陆面数据同化研究.E- mail: shicx@nsmc.cma.gov.cm
  • 基金资助:

    中国科学院知识创新工程重要方向项目(Kzcx2-yw-126-2)、国家863 项目(2007AA12Z144)、公益性行业(气象)科 研专项(GYHY(QX)2007)、中国气象局新技术推广项目(CMATG2008Z04)共同资助.

A Time Downscaling Scheme of Pr ecipitation by Using Geostationary Meteorological Satellite Data

SHI Chunxiang1,2,3, XIE Zhenghui1   

  1. 1. Institute of Atmosphere Physics, Chinese Academy of Science, Beijing 100029, China;
    2. National Satellite Meteorological Center, CMA, Beijing 100081, China;
    3. Graduate School of the Chinese Academy of Science, Beijing 100039, China
  • Received:2008-04-01 Revised:2008-06-01 Online:2008-07-25 Published:2008-07-25

摘要:

高分辨率降水时空分布估计对于陆面数值模拟和同化,天气、气候与环境及其相关领域研究具有重要意 义。长时间序列降水分布常使用时间和空间插值技术进行估计。本文提出了一种基于每小时静止卫星云图云分类 信息的累积降水估计时间加权插值方法,它通过利用高时空分辨率静止卫星云图所反映的云的变化,与卫星降水 估计和地面观测降水相结合,以获得时空分布比较合理的地面降水分布。将该方法应用于FY2C 云图和国家卫星 气象中心FY2C 的6 小时降水估计业务产品,得到FY2C 卫星发射(2005 年)以来具有10KM空间分辨率以及1 小 时分辨率的降水数据集,并利用中国区域自动雨量观测资料对该降水估计数据集进行检验和评估,以表明该方法 的合理性。

关键词: 降水估计, 静止卫星, 时空分布

Abstract:

Precipitation estimation with high quality and highresolution in time and space is significant for land model simulation, land data assimilation, and the weather, climate and environment research. To estimate precipitation distribution from rain gauge, the traditional methods interpolate the precipitation in space and time without considering the function of the cloud and discontinuity of precipitation incident in time and space. The paper develops a highresolution precipitation estimation scheme by using geostationary satellite data, which estimates six- hour resolution precipitation by using multi- channel satellite data and gauge data, and then interpolates the six- hour resolution precipitation into one- hour resolution precipitation by using cloud precipitation probability as weight. The newly developed method is applied to FY2C satellite data and its 6- hour resolution precipitation data to get one- hour resolution precipitation, and is validated again precipitation observation, which shows that the scheme is reasonable.

Key words: precipitation estimation, satellite data, spatial and temporal scale